The ex-Toulouse Judith and Holofernes

UNPUBLISHED MANUSCRIPT / 2019 Setting parameters: The Toulouse Judith and Holofernes Keith Christiansen The object of this article is to suggest parameters for a meaningful discussion about the painting of Judith and Holofernes that was discovered in an attic in Toulouse in 2014 and was sold to a private collector in June 2019. It does not pretend to present new material, but rather attempts systematically to review the documentary and historiographic evidence relating to a lost work of the same theme by Caravaggio as well as to analyze the technical and stylistic characteristics of the picture in question. Because what is at issue with the Toulouse Judith and Holofernes is not merely an interesting picture with a disputed attribution, but a major work of art with evidentiary claim to being by one of the defining masters of European painting. On September 15, 1607—just months after Caravaggio left Naples for Malta—Vincenzo I Gonzaga, who that March had purchased the Death of the Virgin in Rome (Musée du Louvre)—was informed by his agent, Ottavio Gentili, that there was on the market in Naples "qualche cosa di buono di Michelangelo Caravaggio che ha fatto qui" (“something good that Caravaggio has made here”). Ten days later we learn from a missive sent by the painter Frans Pourbus (1569–1622), who was acting in his capacity as expert for the duke, that this "qualche cosa di buono" consisted of "doi quadri bellissimi di mano de M. Da Caravaggio. L'uno è d'un Rosario et era fatto per un'ancona et è grande da 18 palmi et non vogliono manco di 400 ducati; l'altro è un quadro mezzano da camera di mezze figure et è un Oliferno con Giudita, et non dariano a manca di 300 ducati." (“Two very beautiful paintings from the hand of Caravaggio. One is a Rosary that was made as an altarpiece and measures 18 palmi and they want no less than 400 ducats; the other is a mid-sized gallery picture of half-length figures and is of Holofernes with Judith, and they will take no less than 300 ducats”). The documents are conveniently gathered by Stefania Macioce, Michelangelo Merisi da Caravaggio: Documenti, fonti…, Rome, 2010, p. 236 docs. 815-816 The prices asked were considerable: Vincenzo Gonzaga had paid a total of 350 ducats for the Death of the Virgin, and that price included a fee to an intermediary and the costs for packing and transport. See Stephane Loire, Peintures italiennes du XVIIe siècle du musée du Louvre, Musée du Louvre, Paris, 2006, p. 59 For the recently completed Seven Acts of Mercy Caravaggio was paid 400 ducats. The two pictures seen by Pourbus seem already to have already been in the possession of the Flemish painter Louis Finson, who retained partial ownership of them until his death in 1617. It was then, in Amsterdam, following a trail that led from Marseille and Provence to Toulouse, Bordeaux and Paris, that Finson drew up a will leaving his half ownership of these to his partner and colleague, Abraham Vinck, who he had known and worked with in Naples. In his will, Finson leaves to Vinck, “twee stucken schildereyen beyde van Michael Angel Crawats, d’een wesende den Rosarius en d’andere Judith en Holpharnis”: see Louis Bodart, Louis Finson, Brussels, 1970, pp. 228-229, doc. 16. Bodart (pp. 12-16) discusses Finson’s period in Naples and his involvement with these two pictures, based on the known evidence at the time. For a recent summary of Finson’s activity and his involvement copying Caravaggio, see Gert Jans van der Sman, in Caravaggio and the Painters of the North (exhib. cat., Museo Thyssen-Bornemisza, Madrid), 2016, pp. 54-56. (Finson is documented in Naples between 1604/05 and 1612, and Abraham Vinck, who was described as Finson’s “compagno” in 1608, was in the city from 1598/99-1610. In addition to Bodart, op. cit. above in note 3, see, for recent overviews of Finson in Naples and his role as a copiest of Caravaggio, the contributions of Giovanna Capitelli, Antonio Ernesto Denunzio, Giuseppe Porzio, and Maria Cristina Terzaghi, in Giuditta decapita Oloferne: Louis Finson interprete di Caravaggio (exhib. cat., Gallerie di Palazzo Zevallos Stilgiano, Naples), 2013. Also : Paul Huys Janssen and Pierluigi Leone de Castris in Paul Smeets, ed. Louis Finson, The Four Elements (Rob Smeets Old Master Paintings), Milan, 2007, pp. 13-25 and 39-51 ; and, most recently, Maria Cristina Terzaghi, Caravaggio Napoli (exhib. cat., Naples), 2019, pp. 46-50. ) The history of the “Rosario” (the Madonna of the Rosary) can be followed without interruption from its sale shortly after Finson’s death down to its present location in the Kunsthistorisches Museum, Vienna. See Mia Cinotti, Michelangelo Merisi detto il Caravaggio, in I Pittori Bergamaschi: il seicento I, Bergamo, 1983, p. 551; and, for a detailed overview of the complex issues associated with this work, Wolfgang Prohaska and Gudrum Swoboda, Caravaggio und der Internationale Caravaggismus, Sammlungskatalog der Gemaldegalerie : Rome, I, 2010, p. 71. How the picture came to be owned by Finson and Vinck remains a matter of conjecture (but see below, in the text). Equally puzzling are the questions surrounding its patron, destination, date, and the reasons for its appearance on the art market in Naples despite the stylistic evidence that it must have been painted in Rome. In the archives of the convent of the Dominicans in Antwerp is a record of the purchase of the picture following Finson’s death: “e ci fu procurata da diversi amatory, nominatamente dai Signori Rubens, Bruegel, van Bael, Cooymans e diversi altri. Avendo essi visto in quest’opera un’arte straodinariamente grande, e considerando il presso no troppo alto, l’hanno comperata per affetto verso la cappella e per avere in Anversa un’opera così rara, pagandola non più di 1800 fiorini…”. The painting is shown, installed in S. Dominic in Antwerp, in a painting by Peter Neefs dated 1636 (Rijksmuseum); the other altarpieces in the church, such as the Rubens Flagellation, are datable between 1617 and 1620, so it seems probable that the Madonna of the Rosary was acquired shortly after Finson’s death (1617) or that of Vinck (1619). In 1617, Vinck held a sale at which another picture—a copy of the Crucifixion of Saint Andrew, judged in 1619 to be by Caravaggio by the painters Barnaert van Someren, Willemvan den Bundel, Pieter Lastman, and Adriaen van Nieulandt and Louis du Predt—was also sold (another copy was evidently made by Vinck and was with his daughter and her husband, unsold, in 1645). The Rosary was sent as a gift to Emperor Joseph in 1786 and is today in the Kunsthistorisches Museum in Vienna. The second picture—the Judith and Holofernes—disappears after 1617. We do, however, have two notices that may possibly indicate its transfer from the Lowlands to France. The first is the mention in 1689 of a Judith and Holofernes by Caravaggio in the hands of the hugely successful Antwerp printmaker/publisher and dealer in works of art, Alexander Voet. The second is a postmortem inventory of the Parisian collector Francois Quesnel in 1697. For these notices, see Bodart, op. cit. above in note 3, pp. 14-15 note 1; J.-J. Brouchy, Nouvelles archives de l’art francais, 3e s. VIII, 1892, pp. 91-92. The first discussion of these documents and their probable relationship to the Banca Intesa picture is that of Maurizio Marini, “La ‘Giuditta’ del 1607. Un contributo a Caravaggio e a Louis Finson,” in Maurizio Calvesi, ed., L’Ultimo Caravaggio e la cultura artistica a Napoli in Sicilia e a Malta, Siracuse, 1987, pp. 59-82. See also, ibid., Caravaggio: Michelangelo Merisi da Caravaggio: “Pictor praestantissimus”, Newton Compton, Rome, 1987, pp. 503-504 cat. 70. There has been the suspicion that Quesnel’s collection was comprised mostly of copies. Interestingly, Quesnel also owned a Denial of Saint Peter and a Rosary that, like the Judith, were ascribed to the Lombard painter. Whether any of these were originals or whether they were all copies, as has been supposed, cannot be said with any certainty. As it happens, Finson is known to have made a copy of the Madonna of the Rosary that in 1630 was sold to Jacob van Nieulandt through the Middelburg merchant Charles de Coninck, reminding us of the rapidity with which works of art could change hands as well as of Finson’s well documented activity as a copyist of Caravaggio actively involved in the art market (an activity to which we shall return). See Bodart, op. cit. above in note 3, pp. 132-34. Is this possibly the picture cited in Quesnel’s inventory? And what about Quesnel’s Denial of Saint Peter and its possible relationship to the picture now in The Metropolitan Museum? These are questions that cannot be answered with the information currently available. What is clear from these documents is that Caravaggio painted a Judith and Holofernes that was on the market in Naples; that it presumably differed from the picture he had painted for Ottavio Costa and that is now in Palazzo Barberini; that it was admired by skilled appraisers of works of art; and that it was subsequently taken by Finson to Amsterdam, where it was sold following his death. One further fact regarding the picture’s provenance requires mentioning, and that is that when the old relining was removed in 2020, there was found on the reverse of the original canvas an inscription identifying it as from the collection of the Comtesse de Trogoff, née de Flincy. In fact, it appears in her post-mortem sale of over five hundred paintings in 1851, when it was thought to be Spanish. See the 1851 sale catalogue: Hotel des Ventes,rue des Jeuneures no. 42, 7-10 april 1851, no 119. Until 1984, the appearance of this lost work by the great Lombard master was a matter of pure speculation. Indeed, it was even proposed that, contrary to the subject cited in the letters to Vincenzo Gonzaga, perhaps the real subject was Salome with the head of John the Baptist, as two post-Roman pictures of this subject survive (Palacio Real, Madrid and National Gallery, London). The matter is thoroughly discussed in the still essential catalogue entries by, Cinotti, op. cit. above in note 5, pp. 453-454 cat. 26, and 574 cat. 121 However, in 1984 Pierluigi Leone de Castris argued that a painting now in the Banca Intesa Sanpaolo in Naples was likely to record the appearance of Caravaggio’s lost painting. De Castris argued that the Naples picture is by Finson, and while this attribution has been contested, the idea that it, indeed, copies Caravaggio’s lost painting has gained broad acceptance. It is this factor that made the discovery in 2014 of the prime version of that composition something of an event. See Leone De Castris, in Nicola Spinosa, Il patrimonio del Banco di Napoli, 1984, pp. 36-39. His arguments were taken up and expanded by Maurizio Marini, op. cit. above in note 6, 1987, pp. 503-504 cat. 70. For further discussion of the Banca Intesa picture: see Gerlinde Gruber, in Caravaggio e l’Europa: da Caravaggio a Mattia Preti (exhib. cat., Palazzo Reale, Milan), 2005, p. 404, cat. VI.1; and the contributions in “Giuditta decapita Oloferne”: Louis Finson interprete Caravaggio, cited above in note 4. The status of the Naples picture as a copy of the documented, lost work by Caravaggio was forcefully argued by Ferdinando Bologna, “Indice ragionato delle opera del Caravaggio,” in L’Incredulità del Caravaggio, Turin, 1992, ed. 2006, pp. 334-36; see also his entry in Caravaggio : l’ultimo tempo 1606-1610 (exhib. cat., Museo Nazionale di Capodimonte, Naples), 2004, p. 166. Although Bologna did not believe the Banca Intesa copy to be by Finson, he considered its status as a copy of the lost painting “definitiva”. See also Sebastian Schütze, Caravaggio: The Complete Works. Taschen. 2009, pp. 294-95 cat. 82. In his contribution to Giuditta decapita Oloferne, cited above in note 4, p. 13, Gert Jan van der Sman accepted both the attribution of the Naples picture to Finson and the probability of it being a copy of the lost Caravaggio—“per consenso unanime degli studiosi derivate da un’invenzione del Merisi del 1606-07.” He noted shared characteristics with Finson’s signed and dated painting of the Four Elements (Houston Museum of Fine Arts). Howerver, at the study day dedicated to the Toulouse picture that was held in Milan in February 2017, van der Sman revised his view about the Banca Intesa picture and conjectured that the Toulouse painting is by someone close to Caravaggio, though who that might be he left hanging. Prior to the appearance of the Toulouse picture, Gianni Papi, now the primary proponent for ascribing the Toulouse picture to Finson, had been of the opinion that the Banca Intesa picture possibly recorded the composition of Caravaggio’s lost painting: see “Sulla ‘Maddalena a mezza figura’ di Caravaggio”, in Spogliando modelli e alzando lumi: Scritti su Caravaggio e l’ambiente caravaggesco, 2014, p. 64 note 13. Then, in Caravaggio: La Crocifissione di Sant’Andrea Back-Vega, Skira, Milan, 2016, p. 26, he revised that view, noting: “Just because Merisi painted a Judith it does not mean that Finson copied it. Finson’s Judith may have been his own invention, or merely inspired by the prototype painted by his friend and in his possession.” Neither Caravaggio’s documented picture nor the Banca Intesa painting are mentioned by Sybille Erbert-Schifferer, Caravaggio: The Artist and His Work, Los Angeles, 2012 (German edition 2009), which is strange since she discusses Finson’s involvement with the Madonna of the Rosary. For if, prior to the discovery of the Toulouse painting, it would have been possible to argue that the Naples picture was merely a pastiche by an enterprising artist inspired by Caravaggio’s lost canvas (as I, too, was prone to think), this view is no longer tenable. It is now clear that the Banca Intesa painting is a very close copy of a far finer picture. Moreover, that the Toulouse picture is not merely another copy of a still lost original but, instead, the prime version of the composition. This has now been established beyond any possible question, both by the creative technique, which includes evidence of a conspicuous abbozzo visible even to the naked eye—an abbozzo, incidentally, that coincides in all respects with Caravaggio’s practice—and significant changes that were made as work progressed. I was among those who, at an early stage (spring 2015), was invited to study the picture. I was also involved in the organization of a study day held at the Pinacoteca di Brera on February 5, 2017, and authored a summary of the discussions, for which see the a press release reported by Pierluigi Panza, “Il conclave di Brera su Caravaggio”,16 Febraio 2017 in Corriere della Sera: http://fattoadarte.corriere.it/2017/02/16/il-conclave-di-brera-su-caravaggio/?refresh_ce-cp. A full copy of the summary may be found at: https://pinacotecabrera.org/en/dialogo/third-dialogue-caravaggio-readings-and-re-readings/ and is republished in Labarbe, Michel Labarbe, ed., Michelangelo Merisi da Caravaggio: Judith and Holofernes, Paris, 2019, pp. 104-107. For what it is worth, the following are the notes I took during my first examination of the painting on May 8, 2015. “My interest naturally turned to those traits of execution—the handling of the brush; the presence and character of the abbozzo visible beneath the present surface; the evidence of the artist's decisiveness; the blending of colors. All of these hallmarks of Caravaggio's facture—aside from a use of incisions—were readily evident. One can, for example, make out his first intended placement for the fingers of the left hand of Judith's servant Abra by two thickly brushed strokes visible through the surface texture. Throughout the linen sheet on Holofernes's bed one sees Caravaggio's typical handling in the blending of the black to obtain those charcoal shadows that few of his imitators managed to emulate with a like degree of contrast and richness. But no less indicative of his handling are the colpi di penello by which he gave the surface of the bedsheet texture and a quality of light playing over it. The few abbreviated strokes by which Holofernes's eyelids are described seemed to me more startlingly unfinished and unblended than might be expected, but I thought of the Vienna Crowning with Thorns, which has similar bravura passages and is an indication of the speed with which Caravaggio worked on this picture—a speed that nonetheless did not keep him from lingering over certain effects, such as the gleam in Abra's eyes or the stunningly strong modeling of Judith's nose. The directness of the painting is, to my way of thinking, totally characteristic of the artist and there is no possibility that this could be a copy”. There were, however, some things about the composition and figures that bothered me—though not to the exclusion of Caravaggio’s authorship. Among these were the regularity and prominence of the wrinkles on Abra's face and what struck me then as the overly congested part of the composition where the action is described: the knot of hands around the sword. The fact that the pummel of the sword is damascened in shell gold is itself an indication of a prestigious work: the only other pictures I know with gold are the Giustiniani Amor vincit omnia in Berlin, the Odescalchi Conversion of Saint Paul, and the Sleeping Cupid in Florence. These features have become even more evident following the very light cleaning carried out in the fall of 2018 and have been analyzed by Claudio Falcucci, who undertook a technical examination of the picture in 2017 (since then, the picture has been thoroughly cleaned and restored). See Claudio Falcucci, “Scientific Analyses,” in Labarbe, op. cit. above in note 10, pp. 111-119. Nor can there be any question but that we are dealing with a work of exceptional quality and of enormous expressive power. I believe this much can be agreed upon by all who have spent time studying the picture. A notable exception is Jonathan Jones, “Discovery in a Toulouse attic is no Caravaggio,”The Art Newspaper, 311, April 3, 2019. His analysis, predicated on what he sees as the picture’s “ripe assortment of failings and oddities,” takes to an extreme the passionate, highly subjective, responses to the picture that have characterized far too much of the debate regarding its status and attribution. So it is worth emphasizing that Gianni Papi, the leading proponent for an attribution of the Toulouse picture (as well as the Banca Intesa copy) to Louis Finson, nonetheless maintains that it is “sans aucun doute de très haute qualité,” adding, “mais Finson est un grand peintre, et il a connu étroitement le langage de Caravage.” His view has been reported in social media (April 13, 1016) and in the press. For the above citations, see: http://www.lefigaro.fr/arts-expositions/2016/04/14/03015-20160414ARTFIG00034-affaire-caravage-des-doutes-venus-d-italie.php. Important to any analysis is that in its conception the composition seems at once to refer to and in significant respects to move beyond the magnificent picture that Caravaggio painted in the years around 1600-02 for one of his principal Roman patrons, Ottavio Costa (Gallerie Nazionale di Palazzo Barberini, Rome). For a critical review of Costa’s picture as well as a comprehensive analysis of its technical characteristics, see Rossella Vodret, Maria Beatrice De Ruggieri, Carlo Giantomassi and Donatella Zari, in Caravaggio: tecnica e stile, Rosella Vodret, et al, eds., Silvana Editoriale, Milan, 2016, vol.2, pp. 272-301. Once dated to ca.1599—just prior to work in the Contarelli Chapel in San Luigi dei Francesi—the picture has now been related by some scholars to a payment to Caravaggio by Costa in 1602. The pose of the figure of Holofernes is the most obvious link between the two pictures. And yet, whether in terms of the characterization of the protagonists, the concentrated focus of their action and more complex spatial arrangement, or the implied engagement of the viewer, the Toulouse picture represents the story in a dramatically different way. Whereas the compositional procedures of Costa’s painting, with the figures distributed in planar fashion to either side of the central axis, are in conformity with Renaissance notions of balanced contrasts, or contrapposti, in the Toulouse painting all three figures are tightly knotted together within the space created by Holofernes’s tent, the red cloth of which is shown bunched up and raised on one side and draped over a branch at the other. Concerning Costa’s painting, Mina Gregori, in The Age of Caravaggio (exhib. cat., New York and Naples), 1985, p. 257, noted that “the contrast between the servant and Judith plays on the classical tradition in rhetoric and poetry of contrapposto.” She cited by way of example the recommendation found in Comanini’s treatise, Il Figino (of 1591) of juxtaposing a beautiful maiden with an old hag. In most earlier treatments of the theme, Abra was depicted as a young accomplice or as an African servant, as, for example, in the interpretations by Botticelli (Galleria degli Uffizi, Florence), Mantegna (National Gallery of Ireland, Dublin), and Michelangelo (lunette of the Sistine Chapel). Be that as it may, the larger, thematic contrapposto is that between the two women and Holofernes and the contrast between the expression of Judith and that of the Philistine. It is worth noting that Rosella Vodret, in Vodret, ed., Dentro Caravaggio (exhib. cat., Palazzo Reale, Milan), 2017, p. 218, has shown that the contrast between the physiognomies of Judith and Abra extends to the manner of execution: “luminoso e levigato il meraviglioso volto di Giuditta, costruito per fuse velature di colore, evidenti e dissociate le pennellate invece che costruiscono il volto della serva.” The way the firmly resolute—even defiant—gaze of the mature Judith, who is dressed in a widow’s garb rather than the garments of a young seductress, is directed at the viewer introduces a new dynamic—one not dissimilar to the Ecce Homo that appeared on the Madrid market and has been embraced by virtually all specialists of the artist. Terzaghi op. cit. above note 4, p. 49, has characterized Judith’s gaze (bizarrely described as “ammicacante”) and her widow’s garb as “poco comprensibili”. Caravaggio’s interest in engaging the viewer directly and breaking down the psychological space between the depicted action and the viewer is evident in early works such as the so-called Bacchino malato (Galleria Borghese) and the Boy Bitten by a Lizard (National Gallery, London) as well as in his altarpieces of the Entombment, in which the figure holding Christ’s legs turns his head and looks out of the picture, and the Madonna of the Rosary, in which the patron turns his head to engage the viewer. X-rays have revealed that the artist of the Toulouse painting initially directed Judith’s gaze at Holofernes, underscoring the thoughtful reconsideration of these pictorial dynamics. Conceived as a “femme forte”, she is very much the Biblical heroine rather than, in the case of Costa’s painting, a “femme fatale”, at once sexually aroused and repelled by her action. For a carefully argued, gendered interpretation of Caravaggio’s painting—and of the theme in general—see Mary Garrard, Artemisia Gentileschi, Princeton, 1989, pp. 290-336. She writes, “The Judiths of Correggio and Caravaggio are such creatures of masculine fantasy, sexual objects who lack sexual awareness, delivered to the viewer by their worldly-wise maid servants, who are agents of their mistresses’ innocent sexual promise and neutralizers of their independent power.” For Garrard, it was Artemisia who transformed Judith from the deceitful femme fatale to a heroic figure. The Toulouse picture obviously bears on her discussion and it is interesting that in 1951—following the discovery of the picture now in Palazzo Barberini—Longhi had wondered whether Artemisia based her interpretation on the lost Naples picture (Roberto Longhi, “La ‘Giuditta’ nel percorso del Caravaggio,” Paragone, 1951, reprinted in Studi Caravaggeschi, Milan, II, p. 85). More recently, Jesse Locker, Artemisia Gentileschi: The Language of Painting, New Haven, 2015, pp. 68-99, has explored Artemisia’s images of heroic women in the context of the leading Venetian literary academy, the Accademia degli Incogniti. And so far from Holofernes being imagined as the betrayed lover awakened from his drunken stupor and sexual escapade by a sword slicing through his neck, his face expressing horror and surprise, in the Toulouse picture he is portrayed as a repugnantly vicious man, with cruelty inscribed in every feature of his face. His expression is one of excruciating pain and animalistic rage, expressed by his bared teeth, fierce gaze, and the desperate gesture of his spread-finger hands. The old servant Abra is the toothless accomplice whose grotesque, peasant features and prominent goiter create a startling and even shocking contrast to Judith's austere, matronly beauty. Bologna, op. cit., 2006, above in note 9, pp. 335-36, has noted that this depiction of the goitered Abra had a broad impact on other (particularly northern) painters in Naples, and he cites in specific a work by the so-called Master of Pau, now identified with the early phase of Filippo Vitale (see Gianni Papi in Il genio degli anonimi: Maestri caravaggeschi a Roma e a Napoli [exhib. cat., Palazzo Reale, Milan], 2005, pp. 57-58; and, for his identification with the early phase of Filippo Vitale, idem. “Novita per Carlo Selitto”, in Francesco De Luca and Gianni Papi, Davanti al naturale: Contributi sul movimento caravaggesco a Napoli, 2017 pp. 17, 21 note 11, with full bibliography). Bologna further surmised that in painting a goitered old woman, Caravaggio recalled the figure of a male shepherd in Simone Peterzano’s fresco of the Adoration of the Shepherds in the Certosa di Garegnano. And yet, her position clearly identifies her as an active accomplice in the heroic deed, astonished at the resolution of her mistress. Old women are a staple of Caravaggio's art and assume a variety of roles. There is the concerned, matronly figure of Saint Anne in the Madonna dei Palafrenieri (Galleria Borghese); the rapt face of the devout pilgrim in the Madonna di Loreto; the sympathetically pensive innkeeper’s wife in the Supper at Emmaus (Brera, Milan); the goitered witness at the foot of the cross to which Saint Andrew is being bound (Cleveland Museum of Art); the horrified figure who stares in disbelief, her hands clasping her face, in the Decollation of Saint John the Baptist (Cathedral, Valetta, Malta); and the old crone who, jaws clenched, looks on in grotesque fascination as a struggling man gets his tooth pulled (Galleria degli Uffizi). The attribution of the Toothpuller remains as much a matter of debate as the picture under consideration here, but it is a work with an attribution to Caravaggio that extends back to 1637, when it was first inventoried in the Medici collections, and I unhesitantly accept it. See my entry in Andrea Bayer, Painters of Reality (exhib. cat., The Metropolitan Museum of Art), New York, 2004, pp. 175-76; also Mina Gregori, in Gianni Papi, Caravaggio e caravaggeschi a Firenze (exhib. cat., Galleria Palatina / Galleria degli Uffizi), Florence, Florence, 2010, pp. 122-24 cat. 8; and Gert Jan van der Sman, loc. cit. above in note 3, pp. 184-85. From this point on, such figures become a familiar part of the repertory of caravaggesque painters, essential for the roles of procuress or gypsy fortune teller, from Manfredi and Honthorst to Regnier and, most memorably, Georges de La Tour. As already noted, Abra, with her prominent goiter and fixed stare, is shown engaged in a shared mission, whereas in the Costa picture the nut-cracker-jawed old woman plays the comedic role of the astonished accomplice whose ugly visage and exaggerated, bug-eyed expression is drawn from stock characters of popular theater. The knotted fabric of the military tent that defines the space within which the scene of the Toulouse composition is staged echoes the figure of Judith. Again, analogies for the use of a curtain to underscore compositional features go back to Ottavio Costa’s painting of the same subject and are found as well in the Death of the Virgin and the Madonna of the Rosary. But beyond any such formal analogies with Costa’s earlier painting, there is in the Toulouse picture a heightened emphasis on the violent, visceral aspects of the story: a move towards what we might characterize as a baroque staging of the drama, with all that implies concerning the factors of space, the emphasis on dramatic moment, and the active engagement of the viewer. In keeping with this more psychologically complex and engaging approach—a defining trait of Caravaggio’s post-Roman works—light is employed both to underscore the drama and to define the space (note the gradation in the intensity of the light falling on the fabric of the tent at left and its diminution at the right, with the cavity sinking into darkness). And whereas in Costa’s painting all parts of the composition are painted with equal attention and descriptive beauty, in the Toulouse picture there is a notable, destabilizing range, exemplified at its two extremes by the delicate use of shell gold to enhance the decoration on the pummel of Judith’s sword and the boldly (or crudely, depending upon one’s point of view) description of Holofernes’s face and the distinctively concentric wrinkles describing Abra’s furrowed brow. Such variation in the execution of different parts of a picture can be found in some of Caravaggio’s Roman paintings—the Giustiniani Crowning with Thorns (Kunsthistorisches Museum) is one—but increases and, indeed, becomes a hallmark of his post-Roman works. Perhaps the most relevant example is the Rouen Flagellation, in which the nobility of Christ and the cruelty of his torturers is emphasized by the employment of radically contrasting degrees of finish, so that the marvelously realized torso and face of Christ are contrasted with the more summary—in parts brutal—description of the tormentors. The history and dating of the Rouen painting is well described by Gianni Papi in Mina Gregori, Michelangelo Merisi da Caravaggio: come nascono i Capolavori (exhib. cat., Palazzo Pitti, Florence), Electa Milan, 1991, pp. 300-309 cat. 18. Acquired by the museum in 1955 with an attribution to Mattia Preti, the picture was recognized by Roberto Longhi as the original of a composition previously known to him through a copy he had seen in Rome (“Un originale del Caravaggio a Rouen e il problema delle copie caravaggesche,” Paragone, 121, 1960, pp. 23-36, reprinted in loc. cit. above in note 14, pp. 243-254). The autograph status of the Rouen version was for a time disputed with another copy published by Denis Mahon. Both paintings were exhibited side by side on the occasion of the 1985 exhibition The Age of Caravaggio, firmly resolving the matter. They have been displayed together again at the recent, 2019 exhibition at Capodimonte, Caravaggio Napoli: see the entries by, Alessandra Cosmi, in Cristina Terzaghi, op. cit. above in note 4, pp. 134-38, cat. nos. 7 and 8; Terzaghi discusses the issue of Finson and Vink and the production of copies for the active Neapolitan market. If nothing else, these traits help to indicate the pictures by Caravaggio with which the Toulouse painting bears closest analogies: his late-to-post-Roman works. The technical mastery of the Toulouse painting is amply demonstrated by the way the surface of the white linen sheet is enlivened with resolutely applied, impressionistic brush strokes, and the folds of the cloth draped around the Abra’s shoulder are realized by clusters of long, confident sweeps of the brush. Not only do these traits eloquently attest to an elevated quality of execution, they again provide suggestive analogies with Caravaggio’s post-Roman paintings. There are, for example, parallels with the garments of the innkeeper’s wife in Caravaggio’s Supper at Emmaus (Pinacoteca di Brera) and—most conspicuously—with the shirt of the left-hand torturer in the Flagellation in Naples and the figure holding a torch in the background of the Seven Acts of Mercy. If these analogies are accepted, then it seems clear that whoever painted the Toulouse picture was not only aware of Costa’s painting but keenly attuned to the new directions Caravaggio explored in his first Neapolitan period. If the forgoing comments are taken together, there would seem to be ample reason for considering this picture as the long-lost work cited in the documents. Nonetheless, no sooner had it emerged and been shown to a number of specialists, then reservations were expressed and doubts raised by scholars unable to reconcile certain parts of the re-discovered picture with their notion of Caravaggio’s practice. This is most problematically the case with the figures of Abra and Holofernes, where the concentric wrinkles of the old woman’s brow and the coarse description of Holofernes’s face have seemed to many irreconcilable with Caravaggio’s normal practice—and most particularly with his earlier treatment of the theme. In evaluating these features, it should be noted that the painter of the Banca Intesa copy closely transcribed them, whether the straight line of the tyrant’s cheek, the emphasis on the lower lid of his left eye and the prominently highlighted teeth of his savagely open mouth. In other words, they are part of the finished aspect of the picture. What the painter of the Banca Intesa picture seemingly sought to modify was the effect of rawness. Indeed, throughout the Banca Intesa picture we can observe the tendency so typical of a copyist to apply to all parts a veneer of even facture. Given this fundamental difference between prime version and copy, the proposal that they are by the same artist who—so it has been suggested—is to be identified with Louis Finson, seems incomprehensible. The primary proponent that the Naples and Toulouse paintings are both by Finson is Gianni Papi. As remarked above in note 12, his view has been reported in social media (April 13, 1016) and in the press: http://www.lefigaro.fr/arts-expositions/2016/04/14/03015-20160414ARTFIG00034-affaire-caravage-des-doutes-venus-d-italie.php: “Il y a trop d'éléments stylistiques où je ne retrouve pas sa main: dans la tête d'Holopherne (trop chargée, avec ces dents écartées absolument étranges pour l'artiste), dans le visage de la servante (avec des rides excessives, un traitement bien différent que pour celui de la femme assistant à la crucifixion de Saint-André dans un tableau aujourd'hui à Cleveland - même si cela a dû être le même modèle qui a servi pour les deux compositions). Par ailleurs, les mains d'Holopherne ont des reflets de lumière trop nets sur les ongles (ce qui est plus dans la manière de Louis Finson un peintre nordique, grand ami et l'admirateur de Caravage). L'épée me semble exagérément ciselée. La main de Judith qui tient cette arme est au contraire trop simplifiée. De même que son décolleté et sa robe de satin noir (également proche de ce que fait Finson). Le geste de Judith décapitant est sans cette force. Or c'est l'énergie qui caractérise le Caravage: voyez combien est différente la même action dans le Judith et Holopherne du palais Barberini, à Rome.” His view has also been taken up by Terzaghi, op. cit. above in note 4, pp. 48-49. Not only is the Naples picture painted in the blandly conventional and pedantically uninflected manner typical of a copyist, whoever we might think him to have been, but there are absolutely no stylistic features to suggest a common authorship. Quite the contrary. That said, a detailed technical examination of both canvases strongly suggests that the two pictures were painted in close proximity and possibly in the same workshop, and the implications of this evidence needs to be taken into consideration in any discussion concerning their authorship. The close dependence of the Naples copy on the Toulouse picture was first set out by Claudio Falcucci at the study day in Naples in 2017 and is discussed by Rossella Vodret, “Conclusions of the Scientific Analysis,” in Labarbe, op. cit. above in note 10, pp. 121-25 Among the reasons for suggesting Finson as the author of the Banca Intesa painting is what seems to have been a close association with Caravaggio in Naples as well as his documented activity as a copyist, as most clearly evidenced by his two signed copies of Caravaggio’s Magdalene in Ecstasy (Musée des Beaux-Arts of Marseille and a private collection in Saint Remy de Provence, the latter of which is dated 1613). For an overview of Finson in Naples and his role as copyist of Caravaggio, see Giuditta decapita Oloferne: Louis Finson interprete di Caravaggio, cited above in note 4, especially the contribution of Maria Cristina Terzaghi, « Napoli, primo Seicento : Louis Finson copista di Caravaggio », pp. 29-43. See also the discussion of Marini, 1987, loc. cit. above in note 6, pp. 59-80; and Terzaghi, 2019, op. cit. above in note 4, pp. 46-50. These, too, are notable for their pedestrian quality. Yet, given the inevitably anonymous character of the Naples picture, in which the copyist—whoever he was—intentionally subordinated his identity to the task at hand, the more important question is whether Finson—a fascinating yet, in the end, a second-rate artist—could possibly have painted the Toulouse picture. The upper limits of Finson’s capacities are established by what is unquestionably his most original and finest Caravaggesque work, the Four Elements, painted in Naples in 1611 (now in the Museum of Fine Arts, Houston). See the various contributions in Smeets, op. cit., cited above in note 4. In that publication, Janssen, p. 18, cites from a letter written by Nicolas-Claude Fabri de Peiresc on January 13, 1614 recommending Finson with the encomium, “Il a tout la manière de Michel Angelo Caravaggio, et s’est nourry longtemps avec luy”. Despite the unquestionable ambition of the picture and the novelty of the composition, with its whirligig of male and female figures, the style of the Four Elements remains one of merely vulgar realism and the technique conventional, not to say academic. The pictorial emphasis is the same throughout, with none of those enlivening effects obtained by a flick of the brush to suggest the passage of light or delineate the crest of a fold, or that boldness of execution, risking an effect of rawness—rawness rather than mere vulgarity—that characterize the Toulouse Judith and Holofernes. Pierluigi Leone de Castris, op. cit. above in note 9, p. 36, succinctly characterized Finson’s style as “pittura secca e schematica, manieratamente disegnata nei panneggi e sbrigativamente semplificata riguardo alla ricerca sui problemi luminosi.” That the Four Elements is demonstrably by a different and vastly inferior artist lies at the heart of the problem for those seeking an alternative attribution to Caravaggio (indeed, no viable alternative candidate has been proposed). So, to recapitulate: that the Toulouse and Naples pictures could have been painted by the same artist replicating himself really is unsustainable. Indeed, anyone who saw the two pictures juxtaposed in Milan in 2017 must wonder that such a view could be maintained against the overwhelming visual evidence to the contrary—now made even more evident following its cleaning. Second: whether or not Finson painted the Banca Intesa copy, he cannot have painted the Toulouse Judith. Might he—or another painter (such as Finson’s partner Vinck)—nonetheless have had a hand in its execution? At the scholars’ day in Milan in 2017, it was suggested by more than one of the participants that upon arrival in Naples, Caravaggio may have established a joint workshop with Finson—something that makes sense if we remember that the Lombard artist’s refuge in Naples had been unplanned and was certainly not viewed by him as permanent: he always sought to return to Rome. Moreover, the fact that Abraham Vinck—the close associate of Finson as well as a partner in his stock of paintings—was described in a letter of 1673 as "amicissimo di Caravaggio"—is also of obvious interest and further bolsters the possibility of a shared workshop. A fine summary of this matter is given by Terzaghi, op. cit. above in note 4, pp. 46-60. Such an arrangement would have given Finson privileged access to Caravaggio’s work in progress and could explain the technical evidence pointing to the Banca Intesa copy having been painted under special circumstances and probably in the same studio (it is on the same type of canvas as the Toulouse Judith, joined in the same place, and incorporates some of the underlying features of the Toulouse picture). See the comments of in Labarbe, op.cit above in note 10, p. 37. Might we imagine that in such a shared workshop a certain amount of collaboration took place, thereby explaining those problematic and even disturbing features that mark the faces of Abra and Holofernes? This is the position argued by Rosella Vodret in Labarbe, op.cit above in note 10, pp. 121-25. While this possibility cannot be eliminated out of hand, my own feeling is that while it might conceivably be used to explain the regularity of Abra’s furrowed brow, the boldness in the execution of Holofernes’s face is precisely what is lacking in the academic approach typical of Finson’s work (or that of other caravaggesque painters in early seicento Naples). It is, however, something Caravaggio explored with increasing interest and advantage following his flight from Rome, as he realized how his practice of summarily blocking in the salient features of his figures with a rough abbozzo could be used both to achieve a heightened expressivity as well as to enable a more rapid execution. The similarities between the coarse features of Holofernes in the Toulouse picture and the abbozzo of his counterpart in Costa’s painting, as revealed by infrared reflectography, are indicative of this process by which Caravaggio achieved his own understanding of non finito. One need only think of the characterization of the astonished crowd in the Raising of Lazarus—particularly the figure whose face overlaps with Christ’s extended hand—to appreciate how important this non-finito approach became. It is difficult for me to imagine that an artist of the caliber of Caravaggio—presuming him to be the author of the Toulouse picture—would allow someone of inferior gifts to paint a head as important as that of Holofernes while he tended to the beautifully described torso, arms and hands. Nor does it seem to me likely that he would leave the head of Holofernes and the face of Abra unfinished while bringing secondary passages, such as the bed linen and red fabric, to a high degree of completion. To my way of thinking, it makes far more sense to accept the treatment of Holofernes’s face as an extreme experiment in characterization and expressivity through a radical use of the brush—those colpi di penello that seicento critics associated with bravura and mastery. According to Filippo Baldinucci, Vocbolario toscano dell’arte del disegno, Florence, 1681, p. 48: “e dicesi, fatta di colpi quella pittura, la quale l’Artifice condusse, col posare con gran franchezza le tinte al luogo loro, o chiari, o scuri, o mezze tinte o dintorni che si fussero, dando ad essa piuttua un gran rilievo, e facendo in essa apparire una gran bravura e padonanza del pennello e de’colori; tutto il contrario di quelle ptture, che diremmo sfumate, o affaticate.” The IRR of the abbozzo for the head of Holofernes in the Costa painting is illustrated in Vodret, et al., op. cit. above in note 13, p. 294 fig. 18. By these means Caravaggio pushed to the limits the sort of disturbingly violent expression that he had explored earlier in the shield with the head of a screaming Medusa as well as—in its first iteration—the David and Goliath in Museo del Prado (X-rays reveal that Caravaggio first showed Goliath’s severed head glaring out of the picture at the viewer, his mouth open in a scream of anguish, his teeth exposed The x ray of the Prado picture was introduced into the literature by Maurizio Marini, “Un contributo all’iconografia del ‘Davide e Golia’ del Prado”, in Mina Gregori, ed., Come dipingeva il Caravaggio: atti della giornata di studio, Milan, 1996, pp. 135-142, figs. 216, 218. The intensity of Goliath’s stare and his open mouth with bared teeth is an example of an ongoing interest Caravaggio had explored earlier in the shield with the head of Medusa and, later, in the Toulouse Holofernes. This fascination with extreme expressivity was commented upon by Longhi, loc. cit. above in note 14, p. 79): “La brutalità e la ferocia han sempre corso nel mondo e il Caravaggio, quando la esigenza tematica sia così impellente, troverebbe indegno palliarle e smorzarle anche di un punto.”). What a shared workshop would have facilitated was the production of copies of his paintings for an active and avid market—something Caravaggio was familiar with from his first years in Rome, when, according to Mancini, he had made copies of religious pictures for Pandolfo Pucci, the famous “monsignor Insalata”. During the past two decades we have learned a great deal about the market for copies of Caravaggio’s paintings in Rome. Most recently, Riccardo Gandolfi has clarified the role played by Prospero Orsi in promoting—rather than copying—Caravaggio: “Il ‘turcimanno’ del Caravaggio. Propero Orsi tra pittura e mercato nella Roma del Seicento,” in Francesca Parrilla and Matteo Brochia, Le Collezioni degli artisti in Italia: Trasformazioni e continuita di un fenomeno sociale dal Cinquecento al Settecento, Rome, 2019, pp. 85-98. Regarding the proliferation of copies as well as their possible inflation as originals, it is worth recalling that no less a person than the Marchese Sannesio owned a copy of Caravaggio’s Cardsharps that he believed was autograph. The status of the picture and the workings of the market emerge from a trial in 1621 arising from Sannesio’s picture having been lent so that a(nother) copy could be made. Both it and the (second) copy were stolen. From the testimony given by Giulio Mancini, Sannesio’s picture turned out to be only a copy. Mancini, of course, was well informed on the matter since in 1614-15 he had obtained permission to have copies made of some of Cardinal del Monte’s pictures, including the Cardsharps. (Macioce, op. cit. above in note 1, pp. 273-76 docs. 941, 943-46; 950; pp. 283-85 doc. 997). Someone as politically important as Philippe de Béthune, the French ambassador in Rome who knew Caravaggio personally, managed to get hoodwinked into believing that astonishingly mediocre copies of famous works by the Lombard master (now in Loches) were authentic; they were taken to France in 1605 and inventoried as by Caravaggio in 1608! An example of the exceptional demand for works by Caravaggio in Naples emerges from a series of resolutions taken by the governing board of the Monte di Pietà. The governors had received repeated offers to buy the Seven Acts of Mercy at enormous profit. Therefore, in August 1613, they resolved that in consideration of the excellence of the artist, the altarpiece could not ever be sold, regardless of the price offered. That same month they also gave exceptional permission to Don Juan de Tasis y Peralta, Conde de Villamediana, so that he could have a copy made, provided that it be painted by Fabrizio Santafede, Carlo Sellito, or Giovanni Battista Caracciolo—in other words, by a worthy artist. By 1621, the demand for copies had reached the point that the governing board resolved that no one should be permitted to copy the paintings in their church, and in particular Caravaggio’s. For these notices, see Macioce, pp. 271-72, 283 docs. 933-35, 994. We also know that an artist of the rank of Angelo Caroselli, who is documented in Naples between 1615 and 1623, gained a reputation for his copies after Caravaggio that, we are told, deceived experts: see Gianni Papi, “l’Enigma Caroselli,” Artibus et Historiae, no. 65 (XXXIII), 2012, pp. 130-34. It is important to remember that Finson’s two signed copies after Caravaggio’s canvas of the Magdalen and another by Wibrandt de Geest dating from 1620 attesting that the original was by Caravaggio, are only three of a number of copies. Such was the demand for works based on the master’s invention. Whether any of the pictures currently known—including that from the Klein collection—can lay claim to being the lost, autograph version seems to me doubtful. As noted above, Finson’s copying of Caravaggio in response to market demand is well documented. Moreover, the fact that he signed with his own name his two surviving versions of Caravaggio’s lost Magdalene in Ecstasy suggests either an act of plagiarism—passing off Caravaggio’s invention as his own—or an advertisement of its excellence as a faithful copy by someone closely associated with him. He had, after all, earned a reputation for his mastery of Caravaggio’s style, having spent much time with the master. In a letter written in January 1614, Nicolas-Claude Fabri de Peiresc declared that, “Il a toute la manière de Michel Angelo Caravaggio, et s’est nourry longtemps avec luy…”: see Bodart, op. cit. above in note 3, p. 244 doc. 41. For Finson’s work as a copyist, see Keith Sciberras and Antonio Iommelli in Terzaghi, op. cit. above in note 4, pp. 210 and 214 cat. nos. 26 and 17. As we have seen, among the copies he made of other compositions of Caravaggio’s were those after the Madonna of the Rosary and, in all probability, the Crucifixion of Saint Andrew (Cleveland Museum of Art). The latter copyvreplicated the altarpiece that Caravaggio had painted in Naples for the Conde de Benavente, viceroy of the city between 1603 and 1610, and, like the copy after the Judith, it would have to have been painted in Naples. Finson’s picture was sold following his death, and such was its quality that a committee of painters in Antwerp judged it to be an autograph work by Caravaggio. For a discussion of the history of the picture, see Bodart, op. cit. above in note 3, pp. 15, 132-37, 234-236. The picture, owned by the Antwerp painter Pieter de Wit, was sold in 1619 through Jacob van Nieulandt, who in 1630 was to purchase a copy by Finson of the Madonna of the Rosary. De Wit declared that the Crucifixion of Saint Andrew had belonged to Finson and had been sold to Vinck by Finson’s heirs. Two of the same experts—Peter Lastman and Jacob’s brother Adrien, who affirmed that the Madonna of the Rosary sold in 1630 was a copy—declared the Crucifixion to be authentic. Bodart noted that whereas in the case of the Madonna of the Rosary the original was readily accessible, in the case of the Crucifixion of Saint Andrew there was no authentic work against which to compare it. We might also wonder whether judgments have been influenced by Adrien’s relationship to Jacob van Nieulandt and the financial gain to be made. Yet we may question whether Caravaggio would have made a second, autograph version of a prestigious commission from the viceroy—especially at a moment when he was besieged with requests—and that, moreover, this autograph replica passed unnoticed in Naples. But we can understand the confusion—or was it collusion?—of the group of experts in Antwerp who ascribed it to Caravaggio if this copy is identifiable with a canvas formerly in the Back-Vega collection, for that picture is a work of considerable quality. Indeed, in the past it was sometimes proposed by scholars as the original version. Be that as it may, when, in 2017, it was exhibited at the Cleveland Museum of Art alongside the original, its status as a copy was clear to see, and that relationship has been supported by technical evidence. The matter is reviewed by Erin Benay, Exporting Caravaggio : the Crucifixion of Saint Andrew, London, 2017, pp. 105-138; and Richard Spear, ‘Caravaggio’s Crucifixion of St Andrew and the problem of autograph replicas’, The Burlington Magazine, 160 (2018), pp. 454–61. Benay finds the quality of the Back-Vega picture too high for Finson and proposes someone like Caroselli. For his part, Gianni Papi, op. cit. 2016, above in note 9, put forward strong philological reasons for reassessing the Back-Vega picture as a rare, autograph replica of the altarpiece painted for Juan Alonso Pimentel, Conde de Benavente and Viceroy of Naples. After seeing the two pictures juxtaposed in Cleveland, I find it difficult agree with him. Despite minor changes, the picture seems to me to have the character and quality of a copy—probably, but not necessarily by Finson—and not an autograph replica. A real merit of Papi’s study is a compelling reassessment of the involvement of Finson and Vinck in collecting and copying. He also suggests that the expertise of the Crucifixion of Saint Andrew was likely to have been done at the same time as that of Finson’s copy of the Madonna of the Rosary. As was the case with the author of the Banca Intesa Judith, so the artist who painted the copy of the Crucifixion of Saint Andrew must have had direct access to Caravaggio’s original, and this was most probably through a professional relationship and possibly a shared studio with the famously short-tempered Lombard—such as both Finson and Vinck seem to have had. Although we know that Finson and Vinck actively dealt in pictures, it nonetheless seems remarkable that they could have managed to purchase two prestigious works by Caravaggio valued at more than the Duke of Mantua had paid for a rejected altarpiece and equivalent to the cost of the Seven Acts of Mercy. It seems to me far more probable that when Caravaggio departed for Malta, he deposited these two paintings with his northern colleagues to sell. This is also the view of Terzaghi, op.cit. above at note 4, p. 50. But see the comments of Papi, op. cit. above in note 9, p. 18. The Madonna of the Rosary was, as Pourbus noted, a rejected altarpiece that had been returned to the artist, whereas the Judith may, from the outset, have been painted for the market, in which Caravaggio remained invested right to the end. That Caravaggio continued to paint for the market throughout his career, and not merely when getting established in Rome, is demonstrated by the history of the Denial of Saint Peter in The Metropolitan Museum, which in 1613 was owned by the printmaker Luca Ciamberlano, who used it to pay off a debt to Guido Reni: see Michele Nicolaci and Riccardo Gandolfi. "Il Caravaggio di Guido Reni: la 'Negazione di Pietro' tra relazioni artistiche e operazioni finanziarie." Storia dell'arte, 130 (September–December 2011), pp. 47–55. Reni, in turn, presumably sold it to Cardinal Paolo Savelli. Regarding the Madonna of the Rosary, two hypotheses have been advanced concerning its possible patron, who is shown kneeling next to Saint Dominic, turning his head and looking out of the picture. One is that it is the Conde de Benavente—who commissioned the Crucifixion of Saint Andrew. Indeed, known portraits of the viceroy resemble the donor portrait. But why would the Conde de Benavente reject an altarpiece with his portrait in it? Another is that it is the altarpiece commissioned by Niccolò Radolovich, for which the artist was paid 200 ducats in October 1606 (at which time the altarpiece was still “da fare”). The stipulated dimensions agree, but not the description of the subject. The doubling of the price on the open market the following year also seems implausible. And then there is the puzzle that the style of the picture seems to point to its execution in Rome, not Naples. On all these points see Prohaska, op. cit. above in note 5, pp. 76-84; and Papi, op. cit., 2016, above in note 9, pp. 14, 18. It would not be difficult to imagine how, with the Lombard artist’s death in 1610, the two pictures became the joint property of Finson and Vinck. They were then transported to Amsterdam together with Finson’s considerable inventory of marketable works (which included a copy of the Ecstasy of the Magdalene from which he made the further copy in San Remy in 1613 Papi, op. cit.above in note 9, p. 20, makes the suggestion that Finson had managed to obtain Caravaggio’s picture and took it with him when he left Naples. I find this hard to credit, given the competition for works by the artist following his death. Indeed, the competition for available works by Caravaggio makes one wonder why the Madonna of the Rosary and the Judith did not find buyers in Naples. Were they perceived as over-priced? Or is it possible that Finson and Vinck, who until 1610 had possibly been acting on behalf of Caravaggio, decided they would be better off selling the pictures in Amsterdam?). Finson is documented in Marseille in February 1613, Aix en Provence and Arles later that year, Montpellier in July 1614, and then in Toulouse, Bordeaux, and Paris by March 1615. Bodart, op. cit. above in note 3, pp.19-28. I believe that in coming to grips with the polemics that have dominated discussions of the Toulouse Judith and Holofernes—including the suggestion of the possible intervention of a second artist This suggestion was prompted by Claudio Falcucci’s technical examination of the picture, which involved taking a cross-section from one of the furrows of the old woman’s forehead. He noted that “the composition of the uppermost paint layer [of the furrowed brow] is slightly different from the one underneath: both comprise lead white and earth pigments but the fluorescence from the ultraviolet light of the uppermost paint layer is more intense compared to the one just underneath and of an orange color. This is proof that from one coat to another a different binder was used, or that there was a different quality of binder to pigment and that the uppermost paint layer contains bright red pigment that is absent in the one underneath.” This suggested to him the possibility that another artist was involved. It must be said that this is slim evidence on which to base such a radical conclusion, and one suspects that we have a case of an interpretation of technical evidence being inflected to support a conjecture rather than vice versa. There was, for example, no distinct division between the layers, such as a layer of varnish or dirt, and what one is, in effect, describing is two superimposed brushstrokes of different shades and composition, the upper one giving a highlight the lower one. There is nothing unusual in this and taken alone, the evidence does not support the idea of a later intervention, as for example by Finson or Vinck with a view to the sale of the picture after its shipment to Amsterdam. It should further be noted that the recent cleaning has considerably altered the appearance of this area of the picture.—we must confront the unlikelihood that a picture of this quality and distinctiveness, the subject of which matches that of a well documented painting, could be the work of some anonymous follower of Caravaggio. Our knowledge of Neapolitan painting has moved well beyond that point. It is, moreover, worth recalling how many paintings that have now entered into the canon of Caravaggio posed similar problems when they did not conform to expectations based on criteria that proved to be too narrow or inflexible. We might recall that Caravaggio’s unique mural on the ceiling of Cardinal Del Monte’s stilleria in Villa Ludovisi was for many years the subject of seemingly unresolvable disputes, despite the fact that it is described by Bellori rather extensively—albeit with the qualification that its authorship was not absolutely certain (“tiensi...di sua mano”). The problem was that there were no precise comparisons for such a work, with its foreshortenings and shockingly naturalistic treatment of a theme involving figures from Roman mythology. Moreover, conventional wisdom had it that Caravaggio did not paint murals. Mia Cinotti, op. cit. above in note 5, p. 563 cat. 84, gives a fine overview. Even following its publication by Giuliana Zandri, “Un probabile dipinto murale del Caravaggio per il cardinal Del Monte”, Storia dell’arte, I, n. 3, 1969, pp. 338-43, its attribution continued to be doubted by major scholars. There was the assertion of “la inesistenza di veri rapport stilistici”. Further, it was maintained that “i legami plausibili con opera di Antonio Circignani detto il Pomarancio, sconsigliano l’attribuzione”. The reluctance to confront the quality of invention and the idiosyncrasies of its realization are not unlike what has happened with the Toulouse picture—or so it seems to me. M. Fagiolo dell’Arco and Maurizio Marini, “Rassegna degli studi caravaggeschi 1951-1970,” L’Arte, N.S., nn. 11-12, 1970, p. 122, had made the suggestion of Pomarancio—an idea that now seems incomprehensible. One of the very few scholars who continues to reject Caravaggio’s authoriship is, Bologna, op. cit. above in note 9, pp. 347-48 cat. 103, who has bizarrely advance the name of the Mantuan Pietro Fachetti. For a review and technical information, see the accounts of Vodret, De Ruggieri and Gaggi, loc. cit. above in note 13, pp. 160-71. Cleaned, it has become one of the artist’s defining works, and the idea that Antonio Circignani (Pomarancio) might have painted it—something once seriously considered—now seems absurd. Equally, it now seems incomprehensible that as late as 1971, it was possible for a critic as astute as Michael Levey to question the attribution of the Salome with the Head of John the Baptist in the National Gallery, London on the basis of what he perceived to be its striking qualitative deficiencies. Michael Levey, National Gallery Catalogues: The Seveneteenth and Eighteenth Century Italian Schools, London, 1971, pp. 53-55. As a reminder of the importance of considering intention when evaluating variations in style, it is well worth reading Levey’s comments, which are not that different from those of Jonathan Jones, cited above in note 12. The picture was unknown prior to its appearance at auction in France in 1959. His mistake was in applying to the picture—most likely painted for the market—the same meter he used for judging the great, commissioned altarpieces. As a further example, we might recall that both the Martyrdom of Saint Ursula and the Denial of Saint Peter were initially ascribed to Bartolomeo Manfredi. Subsequently, documents emerged confirming their attribution to Caravaggio and they are now benchmarks for our understanding of the artist’s last years (which has led to a revised understanding of the London Salome). The differences between the facture of the Saint Ursula, commissioned by Marcantonio Doria, and the Denial, also indicate the importance of taking into account the destination of a work—whether for an important patron or for the market—when evaluating degrees of finish. See my comments in, “Some thoughts on Caravaggio, the market, style, and chronology”, in Gli amici per Nicola Spinosa, Rome, 2019, pp. 35-44. For some time the attribution to Caravaggio of the Flagellation in Rouen, at one point ascribed to Mattia Preti(!), met with resistance and its relationship with what is now universally recognized as a contemporary copy was disputed. See above, note 19. It’s worth noting that the copy of the Rouen painting modifies the coarse brushwork of the face of one of the tormentors in the same way as the Banca Intesa picture. It too was found to display certain crudities. Doubts plagued the recognition of the extraordinary Saint Francis in Meditation in Cremona (how odd it now seems that Longhi could have it believed to be a copy Cinotti, op. cit., above in note 5, p. 423, gives a fine survey of the uncertainty that still pervaded the scholarship in 1983. As late as 1960 Longhi remained hesitant, awaiting the possible appearance of a stronger version (“piu indefettibile”). Cinotti accepted its autograph status despite the presence of parts she found slack (“Pur nella stesura un poco fiacca di talune parti”). See the recent entry of Mario Marubbi in Rosella Vodret, op. cit. above in note 14, pp. 146-49). Interestingly, the two, last-mentioned paintings display points of comparison with the Toulouse Judith and Holofernes: the furrowed brow of the Saint Francis; the slashing brushstroke defining the lower lid of the eye of one of the figures in the Flagellation. In 1985, it was still necessary to defend the bold, cursory brushwork of the Vienna Crowning with Thorns—another picture presenting certain analogies with the Toulouse painting. The picture is now firmly documented as having belonged to Vincenzo Giustiniani and is universally accepted, the date being still open to dispute as well as the question of whether it was a commission or, as I tend to think, one of the works purchased by the marchese on the market. See the entry of Mina Gregori, op. cit. above in note 14, pp. 316-18, cat. 90. What remains remarkable is that some scholars persisted in seeing the picture as a copy, despite the technical evidence to the contrary. This is of relevance in the case of the Toulouse picture. The loose, summary brushstrokes that characterize the Saint Jerome from Monserrat continue to elicit doubts about the picture’s authorship, though, it, too, is plausibly identifiable with a painting in the Giustiniani collection. For the probable Giustiniani provenance, see Silvia Danesi Squarzina, La collezione Giustiniani, inventari I, Einaudi, Turin, 2003, pp. 111-13, inv. no 40; and the comments by Rosella Vodret and Artur Ramon, in Vodret, op. cit. above in note 14, pp. 132, 225-27. Will we ever reach a consensus regarding the Toothpuller in the Uffizi, which has an illustrious provenance and early critical history and, as I noted as long ago as 1986, displays virtually all the technical features of Caravaggio’s autograph works? See above, note 17, and Keith Christiansen, "Caravaggio and 'l'esempio davanti del naturale'," Art Bulletin, 68, l986, pp. 434-35, 438, where I noted, “Not only is the quality of the highest order, but the actual handling of paint is virtually identical with that found in Caravaggio's late work. If the picture is not by Cara- vaggio, it is by someone who not only imitates his types and compositional ideas, but also employed the idiosyncracies of his working method in a manner indistinguishable from the master's, and on a comparable qualitative level. I unhesitatingly accept the picture as by Caravaggio, and believe that it throws a good deal of light both on the function of incisions in his work and their disappearance after 1605, when the highly individual figures of his Roman pictures are replaced by a store of types. One such ‘type’ recurs in the Toothpuller—the old hag to the right—but the genre nature of the subject and the complexity of the composition would have encouraged Caravaggio to stage the scene with models, as he had earlier in his career.” The list could go on. What is important—indeed crucial—at this juncture, is to recognize what both technical analysis and archival research has demonstrated over the past two decades: that Caravaggio is a far less predictable artist than some would like to think and that a proper understanding of the Toulouse picture cannot be based on a misconstrued and facile comparison with the earlier, very differently conceived and more elegantly painted Judith and Holofernes in the Galleria Nazionale di Palazzo Barberini. We are, after all, dealing with two pictures dating from very different, quite distinct and remarkably complex moments, Costa’s picture belonging to the years in which the artist painted the canvases in the Contarelli and Cerasi Chapels and the Toulouse painting with the Naples Seven Acts of Mercy and Flagellation of Christ. The technical evidence combined with the documentary trail seem to me to establish firm parameters for a discussion of the Toulouse painting, and prior to forming alternative opinions about its authorship, we need continuously to remind ourselves that scholarship can only advance if we are willing to move beyond a too narrowly defined construct of the artist. The errors deriving from such an approach form a conspicuous part of the historiography of Caravaggio connoisseurship. Only by recognizing that there is an overriding case—documentary, technical and stylistic—for believing that the Toulouse painting is the work referred to in 1607 and subsequently taken by Finson to Amsterdam, can we proceed to the next stage of what has proven to be a long learning curve relating to an unpredictable and always surprising artist. ____________________________________________________ 24